Readings in the Theory of Income Distribution [2nd (1949) ed.]

Readings in the theory of Income Distribution, published by the American Economic Association

331 25 11MB

English Pages 365 Year 1949

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Readings in the Theory of Income Distribution [2nd (1949) ed.]

Citation preview

READINGS IN THE THEORY OF

INCOME DISTRIBUTION

Selected b y a C o m m i t t e e of THE

AMERICAN

ECONOMIC

Ptiladelphia

ASSOCIATION

Toronto

THE BLAKIST C^^ COxMPANY 1949

C.'PVRiGHi,

1 9 4 6 . U Y T H E BL,\KISTON

1 \ ' S "ook isS:,y :t

prot', Vt,- by cejyright,

titih the f^ctilir-n

of shcrl

RfprintJ

oUfS'TE'i E\

!>' T H E i r - i l T E D STAT

.rlE MAPLE

PRESS

and no f-.^.tl -J

lui: itior-,

Aui^u :

COMPANV

for

rr't,tij\

7949

Of

\»flO':\

YvlRK, PA,

PREFACE T h e Blakiston Series o f Republished Articles o n Economics designed to make accessible in its successive volumes the most us^ ful articles and essays in the various fields of economic analysis a n . policy. By an agreement between T h e Blakiston Company s. the American Economic Association, the Company, during o initial experimental term, will undertake the publication of \ annual volume devoted to a subject chosen Ijy a standing c o m m i t t C i of the Association and edited by a special committee of experts o i that subject. T h e fav'orable reception o f the first two volumes i i the series has encouraged the publishers to continue the ventun despite technical and editorial dil^culties entailed by the war a n reconversion periods. Primarily, t h e Series is oriented toward the tuition of senior an graduate university students; but there is also the hope that pro''° sional economists will find it a useful means of keeping abreai developments in fields other than their own. T h u s the Series m: help to lessen the intellectual provincialism of specialists, whic is said, threatens them with having nothing in common— not ev a n education. Furthermore, despite the aridity which economi; s e e m to cherish, and despite their affection of technical jargon a r even of terminological monsters called u p for the occasion, intelligent layman will in general be' able to find his way success, fully tlirough t h e s e collections, to his lasting benefit if not dclig' It should go without saying that the editorial committees de.u not only t o acquaint the reader with doctrines common to ma o r all o f the scholars in the field, with the contentious issues, a with the m.ore interesting idiosyncrasies of certain writers, but . t o d o all this without bias in the selection o f content. Editorial responsibility for the present volume was b o r r Professor Bernard Haley until the pressure o f duties in the D
' .19.

W A C E P o l ' J I E S OF BY JOHN 1 .

T^ADt,

33f

LiNioN::.

U;..Lt>

F r o m Arr-r/i.-un

I'.snomir

Rsvieit.,

1942

;^

INTEREST 20.

T H E M v T H O I OGY CK C A P I T A L B Y F R - F D M C H A . v.

F r o m Qu-iierly Jiornd 21.

C A P I T A L AND

355

H.WSK.

of t.,onsmics,

1935 -1936 384

iNTKREST

B V I'RANK i:!. K -:::i!r F r o m Ftyaoj-r-iia

22.

Brittanka,

1946

T H E T H E O R Y OF THE R I T E OF INTEREST

418

B Y JOHN MAYNARLI KEYNES

^

F r o m The Lessonsoj Monetary Experience, Essays in Honor oJ |«23.

-n^ F:ju:,

iy3/

M R . K E \ N E S AND T)rE R A T E OF INTEREST

'425

B Y D E N N I S H . P I-JBERTSON

From Essays in Monetan 24.

Theory, 1940

M R . K E Y N E S AND TITE " C L A S S I C S " ,

SroGESTED

INTER-

PRETATION

461

B Y JOHN R . HICKS

l-rom Econometrica, 1937 ' 25.

MoNE'iARY

POLICY AND T H E T H E O R Y OF LSTEREST

.

.

.

477

•.

499

BY H A R O L D M , S o w E i ?

F i o m Quartrrly Journal oJ Economics, 1 9 4 0 - 1 9 4 1 26.

T H E STRUCTURE OF INTEREST R A T E S B Y FHIEDRICH

LUTZ

From Quarleny Journal of Economics, 1 9 4 0 - 1 9 4 1 PROFIT ^'21.

PROFIT B Y FRASK H

533 KNIGHT

From Lncyclofufdia

nf thi S'vrial Snenc'.s, '.934

yv^i 28.

CONTENTS RISK, UNCERTAIN n - , AND THE POU.NDiNo B Y Ai BEni

Fruin 29.

L'NPROFITABILITY OF C O M -

PROBABIIITIES

547

/%IIORI'HART

Sidi-.fs in M^ir.tm.Uirid

EMERPRISF,

PR^iFTli.

:ono'nks end Economibxr.i,

AN'D TH?, M o D E R N

1942

CORPORATION^

-.58

B Y Roi-'RT. A. G O R D O N ,m E-:ploralior^ in Economics, ' " 3 0 30.

CoRFOR.vrE

KARM\(;S

os

INVESTED C A P I T A L

571

BY' W i i LEONARD r - R i ! « Fro.Ti Hjfiard Bi^sintK! Hf.ifw, 1 9 3 8 R E N 31.

r

T H E i T t s i r s r c A L A P P R O A C I ! TO l rent 1913 'dollars) dollars)

Total Income

(m

Tear

t1,000,000)

1913 1911

35,723 2,000

368 278

368 296

1928 1928

89,419 5,938

749 604

541

1911 1913 1913 1913 1914 1913 1913 1"13 1914 1913 1913 1900-14 1913-14

9,840 11,934 6,387 1,251 3,659 660 891 1,366 2,149 7,216 1,156 3.597 1,473

234 178 161 164 102 17! 132 64 105 52 22 12 300

250 178 161 164 108 171 132 64 94 52 22 14 292

1928 1928 1928 1924 1928 1924 1927

18,730 17,990 7,856 1,438 4,944 1,131 940 764 3,497 16,434 5,492 6,496 3,165

411 279 192 187 121 289 141 87 162 107 89 20 504

293 199 188 135 96 178 152 85 117 62 53 13 304

1929 1923 1929 1925 1921-22 1927-28'

401

* T h e total i n c o - n c o f e a c h c o u n t r y w a s c o n v e r t e d into A m e r i c a n d o l l a r s o n the basis o f t h e a v e r a g e r a t e o f e x c b a n g r o f i^e- c o r r e s p o n d i n g y e a r a n d d i v i d e d by t h e p o p u l a t i o n ftg-irc t o g i v e t h e p e r c a p i t a i n c o m e i n current dollars.

T h e t o t a l i n c o m e oi e a c h c o u n t r y , e x p r e s s e d in its own c u r r e n c y , w a s d e f l a t e d b y t h e -.vholc-

s a i c price i n d e x for t h a t c o u n t r y w i t h 1 9 1 3 as t h e b a s e y e a r , c o n v e r t e d i n t o A m e r i c a n d o l l a r s o n t h e b a s i s o f t h e r a t e o f e x c h a n g e i n 1913 a n d divided by tlie p o p u l a : i o n f.g^ire to g i v e t h e p e r c a p i t a i n c o m e i n 1 9 1 3 d o l l a r s . T h e quoiatioM o f exch:int;r r a t e s u s e d a r e u k c p from J . R . M o o d . " H a n d b o o k o f F o r e i g n C u r r e n c y a n d E x c h a n g r , " U n i t e d Stages, B u r e a u of F o r e i g n a n d

Dome5tiC C o m m e r c e , Traie

Serin,

pTomotion

n o . 102 ( 1 9 3 0 ) .

W h o l e s a l e p r i c e i n d i c e s a r e t a k e n trorn VV- C . M i t c h e l l , " I n d e x N u m b e r s o f W h o l e s a l e P r i c e s , " U n i t e d S t a t e s , n o . 284 (1921) a n d from C a n a d a , D o m i n i o n Bureau of Siatistics,

Bureai.i u f L a b o r S t a t i s n c s , Bulletin, and F'lce

IndiJ^ts,

Prices

1^13-1931 (Ottawa 1932).

t E x r i u d c s incomf ih^ Wealth

la si.i^iltsation

Foi B e l g m m : Bau-

f. ses cunsiquences

(2nd -d.

and .Vc.ia?ii/ ^ f o m ? of Several

-je/ore and ajler the ft'ar ( R o m e 1 9 2 3 ) ; M c l i a d o , L . . " 1 1 rcLldi'o p r i v a ^ o d c g l t haH-^ni n e l 1 9 2 8 " i n

voi. ix, n o . 3 - 4 (1932) 2 5 1 - 3 2 1 ,

Important Afetron,

F o r S w i u e r l a n d : M o r i , P . , " D a s K b w e u c r i s c h e V o l k a e i n k c m m e Q " i n Z*^'-

NATIONAL INCOME

tant criterion of economic welfare, and, if the series were long enough, would suggest whether the nation tended in the course of time to grow richer or poorer and how rapidly the change was taking place. Estimates of total income are also employed in ascertaining the proportions in which it is or may be divided among social classes, between the community and the individual, between consumption and capital accumulation, and the like. However used, figures like those given in Table I appear to be quite serviceable; they seem to measure in comparable units something quite definite and significant. Further investigation reveals, however, that the clear and unequivocal character of such estimates is deceptive. Theoretical problems airise in defining the area of " n a t i o n " ; in the choice of stage in the circulation of commodities and services at which income is to be segregated and measured; in the inclusion, exclusion and basis of evaluation of various commodities and services that are to be added into a national total. Finally, variations among estimates may arise from differences in the types of statistical data used and methods employed. Problems in regard to the area covered by national income estimates are due to differences in the location of productive agencies and in the political allegiance and place of residence of their owners. We may thus distinguish productive ageiu' , located within the boundaries of the given state (.4) from those located outside ( 5 ) ; and among them, those owned by subjects of the state residing within its boundaries (aA and aB) or outside JFA and bB) from those owned by aliens residing within the given uhri/l

fir

t!>twn(,trutU

Slalistik,

v o l . Ixii ( 1 9 2 6 ) 5 1 2 - 4 2 .

F o r .Austria: H e r t z , F . O . , " Z a h l u n g s b l l a n i u a d

L e b e t u f a h i g k c i t O s t c r r c i c h s , " V c r e i n fiir S a j i a l p o l i l l k , Scfiriftrn, Kapiioltiildung,

u n j Volkstinkommcn

in Oiterreich

(\'ienna 1 9 2 8 \

k o t n r n r n C^^-rrcJchs u n d Ungarn.^'' in Statistisclu

vol. clxvii ( M u n i c h 1 9 2 5 ) , a n d

n.a., v o l . x x i ( 1 9 1 6 ) 4 8 5 - 6 2 5 , a n d "\^-z rc^*c^u

n a t i o n a l d c la H o n g r i e a c t u c l l c " i n I n s t i t u t I n t e r n a t i o n a l d c S t a t i s t i q u e , Bulletin, 455.

K^pitalMarf,

F o r H u n g a r y ; F r l i n c r , F . von,**Da5 \ o l k s r m -

M^nnt'Srhrifty

v o l x x v . n o . 3 (l*).^!;! 3 6 7 -

F o r S p a i n ; V a n d e l l o s , J. A . , " L a n c h e s s c c t l e r c v c n u d c la p i f n i n s u l c i b c r i q u e " i n Mftron.

no. 4 (1925) 1 5 1 - 8 6 .

,A

v

F o r R u w i a : K a t s , V l a d i m i r , .Varorfru < / o t W . y W R i jjii m r f - ^ i f / d i ; ( N a t i o n a l i n c o m e o l

t h e l ' . . S . S . R . a n d 'tis di-jtribution) (.Vfoscow 1 9 3 2 \

F o r J a p a n : M o n , K . , " T h e E s i i m a T c of ' h r N a t i o n a l

W e a l t h a n d I n c o m e of Japan P r o p e i " in I n s n t u t I n t r r n a t i o n a l d c S t . t t i s t i q u e , Bui.Vfin, v o l . x.>v, pt. ii y T o k y o 19?n

p. I 7 9 - 2 U 4 .

(Bombay

1924).

For I n d i a : S h a h , K .

T . , a i d K . h a m b a l a . K . J . . }V,aIlh

F o r A u M r a h a : S u t c l i l l e , J. T . , Tlu Nalinnal

Diiidmd

and Taxubl,

C^!l JnJia

( M e l b o u r n e 1 9 2 6 ) : W.x>d, G

L.,

" S u r v e y of P r o d u c t i o n a n d t h e N a t i o n a l I n c o m e " i n A m e r i c a n A c a d e m y o f P o l i t i c a l a n d S o c i a l S c i e n c e , AuuU,

vol. d v i u ( I 9 3 t ) 2 6 - 3 0 .

6

CONCEPT OF

OME AND DISTRIBUTION

state (cA and cB) or residing outside {dA and dB). T h e strictly political definition of the area of national income would then include aA, aB, bA a n d bB, and the purely territorial definiuon would include aA, bA, cA and dA. T h e definition used prevalently in national income estimates does not follow either the strict political or the territorial principle but that of residence of the owner of producti%'e agencies, thus including aA, aB, cA and cB, This definition conceives a nation as basically a group of residents within state boundaries and thus anchors national income to the material base of the economic system whose product is being measured. It is preferable, from the point of view of economic analysis, to the purely political definition of the area. But it departs from the territorial principle by including in national income receipts from capital invested abroad, by excluding payments on foreign capital functioning in the country and by counting into national income the earnings of shipping and other internationally operating agencies owned by the country's residents. The property principle is thus allowed to cut across the territorial base, implying the existence of normal conditions in the sphere of international trade and capital movements. There remains a question as to the serviceability in economic aivalysis of area units, defined by state boundaries and supplemented areas of net foreign investment and activity. In a political ci.uity that happens to be a comparatively independent economic system, such as France, national income does measure the combined effect of related and integrated economic forces; b u t for economically dependent political entities, for instance, post-war Austria, national income ought to be studied in conjunction with the incomes of countries closely related; and a similar procedure appears advisable for a country with politically independent hinterlands. For many purposes the political unit approach, so prevalent in current statistical procedure, ought to be supplemented by a breakdown of national totals for some countries and a combination of nadonal totals for other countries. The preliminary definitions of national income given above distingtiish income produced, received, consumed and enjoyed. T h e

NATIONAL INCOME

/

first three run in terms of commodities and services which arc separable from the individual agents and capable of measurement in common units because of the leveling process of market valuadon. T h e last runs in terms of subjective feelings, whose commensurability for various individuals is to be doubted and whose relation to the objectively perceptible economic goods is not, in the present state of knowledge, determined with sufficient precision to permit even purely qualitative economic analysis. Consequently the concept of income enjoyed has to be abandoned in favor of such cruder approximations as income received or consumed. T h e consideration of the level of subjective feelings is not, however, completely omitted; it is retained as a background of the analysis of n.itional income at the measurable stages of circulation of goods and services and dictates some of the methods of detailed analysis. As between income produced, received or consumed, the choice on theoretical grounds depends upon the function which the income concept is expected to perform, cither as a summary or as a n appraisal notion. If the national income concept is a summary of the play ofeconom.it forces, the choice depends upon an analysis of these forces indicating \vheiher production, distribution or consumption is the stage at which the combined effects of the factors analyzed appear most clearly. And if income is an appraisal concept, the choice depends upon the basis of the appraisal: economic power as reflected in total productivity (income produced), individual'' potential welfare as expressed in the purchasing power of the incomes received or individuals' direct welfare as reflected in incomes consumed. I n addition there is another, subordinate basis of choice: the quantitative definiteness of the national income total as revealed by statistical practise. I n such practise the attempt is to provide a measure that could satisfy more t h a n one purpose; and income produced, as the concept of the widest reference, seems at first most suitable. From this one can, by .segregating savings of business units, measure income received by individua's and, by further subtr.-^.ction of individual savings, obtain income consumed. But the shift from income produced to income received involves an estimate of savings of business units, which is

8

CONCEPT OF INCOME AND D I S T R I B i m O N

clearly affected by the accounting procedure, may yield widely varying results in different times or different countries and is not likely at any given time or in any given country to reflect faithfully the value of commodities and services produced but not m a d e available to individuals. Similarly, the further shift from income received to that consumed meets an obstacle in durable goods, for which the estimate of consumption for any given year can be m a d e only on the basis of forecasting the future from the past life history of such commodities. These difficulties are more than statistical: they indicate that in current reality the most clear cut, general concept of national income is income received by individuals; and that the uninterrupted flow of commodities and services through the economic system is best arrested for the purpose of analysis and measurement at the point when the stream reaches the living individuals, after it leaves the productive units proper and before it has been diverted into the vairious channels of consumption. T h e discussion below of the contents of national income is carried on primarily in reference to income as received by individuals. The inclusion, exclusion and evaluation of commodities and services that are to be added into a national total offer the widest range of theoretical problems. The modern economic system consists oi individual units whose basic purpose is making a living; of purely business units whose main aim is the making of profits; and of social organizations whose primary purpose is to render service to society as a whole. Each of these groups contributes differently to the sum total of commodities a n d services produced and distributed, is motivated by forces of quite different nature and involves a different approach to income as an appraisal concept. In face of such diversity national income must be a single entity, reflecting the contributions of these various types of units and reducing them to a common unit of measurement. Since in the modern economic system it is the market, through its e.xchange of goods for money, that provides such a unifying mechanism, it is natural at first to identify national income with the sum total of money payments flowing to the individuals from the market. But uch as simple definition immediately suggests a number of questions,

NATIONAL INCOME

^

arising from a conception of economic goods as having an existence and value independent of the changing market. These questions fall into three broad divisions; first, those reladng to commodities and services to which there is no corresponding flow of money payments from the market to the individual; secondly, those relating to the exclusion of money payments to which no commodity or service corresponds, or to the adjustment of the money flow to reflect more properly the volume of economic goods involved; and, finally, those relating to the distinction between gross and net income. Commodities and sen.'ices to which no flow of money payments corresponds may be divided into three groups. T h e first consists of goods and services received in barter (as over against money exchange), such as farm rents paid in kind, food and board of farm workers, food, board and clothing of soldiers, sailors and all employees whose subsistence in whole or in part is supplied by the employer. From the point of view of the nation's productivity or welfare the omission of such bartered goods would obviously understate the total performance of the economic system. T h e second group consists of goods and services received gratis. T h e difficulty here is not the lack of monetary form but the absence of any specific productive service rendered by the recipient. In such cases, since no production of new economic goods takes place, it appears advisable to exclude the goods from the national income total. If an individual receives charity or a gift this is but a loss on the part of the donor (whose income has been recorded fully elsewhere); and to count the incom.es of both donor and recipient involves either double counting or the consideration of the charity or gift recipient as a producer of service to the donor, an obviously far fetched conception. T h e problem becomes more complicated when such free flow of goods (or money) is directed not from individuals but from the business system, either directly or through such social agencies as the government or charitable foundations. Such free goods, whether in form of money or of commodities and services, must obviously be counted in somewhere in the national total. While their statistical estimate is difficult, their analytical

10

CONCEPT OF INCOME AND DISTRIBUTION

a n d quantitative importance is appreciable and likely to grovf in the future. T h e third type of commodities and services for which there is no corresponding money payment comprises those produced and consumed within the individual economic unit. Here the main problem lies in the segregation of economic from noneconomic activity, since only a rigid line between the two will enable one to include in or exclude from national income such items as commodities produced as a hobby, services of durable goods used in the household or personal services of housewives a n d other members of the family. But there is no h a r d and fast rule by which economic activity can be distinguished from social and individual life in general. T h e importance of economic motives, the regularity of the activity, the relative proportion in which the resulting commodities and services usually appear on the market— all have to be considered. For the modern western economic system no doubt appears as to the propriety of including in national income commodities regularly produced and consumed within the household when they form a part of a larger total destined for the market, most conspicuously, for instance, the share of the farmers' produce retained for their own consumption. Similarly, the estimate of national income should include net services from houses owned and inhabited; but there is considerable doubt as to the propriety of including net services of other durable goods, such as automobiles, furniture or clothing; and more doubtful still is the inclusion of hobby products, for which moreover no statistical estimate of value can be made. Finally, there is general agreement among students of the problem as to the exclusion of housewives' services and services of other members of the family, in spite of the very large size of the items involved. It is recognized that these activities are motivated very largely by non-economic considerations and form much more a part of life in general than of professional economic activity proper. Being conditioned by the institutional set up of the family and of economic society, the line between economic a n d non-economic activity shifts from country to country and from time to time. T h e statistical investigator can lend formal precision to his definition

NATIONAL INCOME

11

and measurement by extending the area of economic activity all the way through the individual household and thus abolishing noneconomic areas. H e will then estimate the values of all personal services rendered within the household and avoid the paradox, mentioned by Pigou, of a decline in national income resuking from a man marrying his housekeeper or that of the increase of national income because of the shift of women into industry during the World War. Such income concepts, however, achieve consistency at the expense of reality, for they disguise the basic, if shifting, line of difference between economic and non-economic areas existing in real life. T h e alternaUve is to adopt a narrower definition of economic activity and allow it to shift, the resulting income measurement reflecting changes in time or differences in space in the performance of a n economic system with a slowly or rapidly shifting area. T h e exigencies of statistical work render the acceptance of the narrower definition of national income almost inevitable, b u t without a recognition of the limitations involved in its use any comparisons of national income across wide dme intervals or between countries marked by essential differences in the relative scope of the economic system are highly misleading. If not all commodities a n d services produced (or m a d e available to individuals) are reflected in money flow from the market to individuals, neither are all money payments to individuals a bona fide reflection of economic goods produced or distributed. Such identity is lacking first in the case of gratuitous money payments, i.e. gifts, charity and relief. Like the similar flow of c o m m o ^ t i e s and services, these should be excluded from national income totals except in so far as they constitute indirect flow from the business system to the individuals and have therefore not been recorded elsewhere. Quite analogous is the second group of cases, those in which money payments come before or after the service is rendered; the essential element in both groups is the fact of a money flow to individuals without a corresponding return within a given period of time. Hence pensions, disbursements by business enterprises out of surplus or capual, withdrawals of accumulated interest on savings by individuals, all may be treated in the same way as charity or

12

CONCEPT OF INCOME AND DISTRIBUTION

relief. I n so far as these payments flow directly or indirectly from the business system and have been deducted as expenses in arriving at net income, they should as a pau-t of the national income produced be included at the point of receipt. I n so far as they are paid by the individuals themselves out of their past or present shares, from which no deduction has been or is made, the counting in of such payments constitutes duplication. If for a given year the accumulated but not withdrawn interest on savings is counted into individual income, it cannot be counted again when such accumulated interest is withdrawn. T h e principle has a similar application in cases of such savings schemes as insurance or building and loan funds. A m u c h wider range of questions arises with a departure from the market valuation scheme as such, of which the first step is a n attempt to adjust for changes in the monetary unit itself (statistical deflation). Of the numerous and intricate problems raised by this procedure only one need be mentioned here: the impossibility o f getting price quotations for qualitatively uniform commodities over a range of time or from country to country. Deflations, while necessary, are consecjuently at best crude approximations; they are the less reliable the longer the span of time they cover or the larger the differences in tlie countries compared. Two further groups of amendments to the market scheme are suggested by economic theory: one based on distinguishing economic activities by the material nature of the results, the other distinguishing them by their organizational character. T h e first relates activity to wealth. Since wealth was originally conceived in material terms, only that labor was considered " p r o d u c t i v e " which resulted in material goods. This definition excluded from productive activity all services and considered incomes from services not as primary but as derived shares. While this point of view has proved inadequate, the distinction persists as a background to much current thought and retains its significance as a contrast which would be valid were our economic system t o become again an economy o f want rather t h a n one o f surplus. I n such eventuality

NATIONAL INCOME

13

commodities as a whole, in so far as they form directly the support of life, would be more important than most services. T h e principle of material productivity survives also in the still current distinction between productive and unproductive loans of capital. Capital invested in manufacturing, mining, trade a n d similar occupations is supposed to be productive because it facilitates the adoption of more efficient methods of production a n d thus results in a greater excess of product over outlay. Hence interest payments on such capital investment represent bona fide incomes and are to be included in national income. But loans to Consuming bodies, whether individuals or public, are unproductive since they are used for direct consumption, a utilization which by its very nature appears incapable of yielding an economically measurable surplus. Hence interest payments on such uses of capital are only a draft upon bona fide incomes. This distinction, hov/ever, seems to arise from a failure to carry through logically the whole treatment of interest on capital and is, in the case of individual consumers, contingent upon the acceptance of some form of tlie iron law of wages. Consumers are generally also producers, and loans to them may serve to raise or preserve earning capacity. -Additional objections are made by many students to the inclusion in national income of interest payments on loans to a public body, like the government, because of the further doubt as to the productive character of governmental activity in general. But so far as governmental activity preserves and raises the productive cheuacter of the economic system, interest payments on government loans are of the same economic nature as interest payments on privately invested capital. T h e current paradox that an increase in government loans would, if payments on such loans are included in national income, serve to raise national income presents no puzzling aspects if it is realized that a rise in indebtedness of private industry would similarly raise the volume of national income. At present a distinction more significant than that based on material productivity is the one between acti^'ities who^ income yielding power is conditioned by the present organization of the

14

CONCEPT OF INCOME AND DISTRIBUTION

economic system and those whose income yielding power would b e retained or even augmented under a different social organization. T h e problem of services appears again here. T h e valuation of personal services .sold on the free market is dependent upon the e.xisting personal distribution of income—as manifested in the contrast between the emoluments of those who cater to the richer classes and the low rates at which a number of services are rendered to small income recipients. Should services be included at the high (or the low) prices which they fetch because of existing income distribution a n d thus be allowed to distort national income totals? T h e same question applies to commodities in which quality distinctions permit different pricing for various groups of income recipients. T h e problem may be generalized by recognizing that the extremely high or low valuation of some commodities a n d ser^'lces is but a partial case of monopoly incomes, whether on the demand or on the supply side. Each investigators economic philosophy vvill inliuence him either to acquiesce tacitly in the valuation within the current economic organization or to attempt some correction for its distorting influence. O n e might correct for the differences in valuation of the same commodities and services among tlie various income groups, just as one corrects for changes over time in the prices fT identical commodities and services. O n e income group could then be adopted as a basic one (just as in the other comparisons .som.e one year is taken as a base), and all commodities a n d services produced could be revalued at prices charged to this basic income group. The practical statistical difficulties of any such correction, however, are enormous, and it is rarely undertaken. Consequently a comparison of the absolute volume o f national income among countries which differ greatly in the personal distribution of income and in the presence of monopolies, of so-called friction incomes, such as advertising, is likely to b e misleading. Even for any one country comparisons of deflated income totals over a period of time are usually dangerous, in so far as the available price data reflect less adequately than d o the income totals the change in the monopolistic areas of national economy.

NATIONAL INCOME

15

T h e monopolistic aspect of some income categories assumes a particularly interesting form in the case of government, where the problem arises as to the valuation to be applied to services thus taken out of the jirea of free play of economic forces. T h e question whether government services are paid at prices warranted by the free play of forces in a competitive market is at bottom unanswerable, for the simple reason that if such play of free forces yielded a n effective measure of those services the latter would probably have been left to private initiative. Consequently market valuation of government services as a whole cannot be made, not even for separate groups of services, except for some highly specific units. It can be estabhshed whether the compensation of a governrhent stenographer or postal employee is higher or lower than that of a similarly trained and employed person in private service, b u t for larger groups of government services it is not the free market b u t the court of enlightened public opinion that can pass j u d g m e n t as to the presence or absence of excessive compensation. I n statistical practise in the United States and most other countries incomes paid to government employees are included in national income, with the recognition that their monetary value is the only available, while admittedly rough, approximation to the value of services these employees render; similarly, pensions for past services and interest payments on loans are included as the equivalent of past and present economic goods produced. Taxes paid by business units are deducted in arriving at the net income of these units, just as are all other business expenses. But taxes paid by individuals are not subtracted from individual incomes, on the assumption that the valu^ of government services to individuals is equivalent to the amount of taxes which they pay and should thus be treated in the same fashion as the individuals' expenses on food, clothing and shektr. This assumption, as indicated above, m a y not be strictly true; and the resulting free incomes or losses to individuals (flowing from the business system via the government) should be included in national income. There is, finally, the possibility of a complete a b a n d o n m e n t of economic bases of valuation a n d the substitution of norms derived

16

CONCEPT OF INCOME AND D I S T R I B i m O N

from other sciences. National income and individual incomes might be expressed in energy units, labor hour units or other standards. Theoretically feasible as such shifts may be, they are as yet not sufficiently developed to merit extended consideration. O n e distinction, however, is of some importance—that betvveen wantability and usefulness. For it raises the question as to the inclusion in national income of the harmful commodities, like opium, and of other commodities or services that appear completely useless from the point of view of a physically and psychologically normal individual. Some practical aspects of this question are predetermined by the fact that with legal prohibition of certain commodities and services an adequate quantitative determination of their volume becomes almost impossible. O n the other hand, data of income tax statistics often include receipts from illegal activities. Theoretically an attempt to make national income a gauge of scientifically determined, real welfare of the population involves an unwarranted optimism as to the validity of sciences concerning h u m a n nature. From another point of view, that of the nation's productivity, it should be recognized that the diversion of a certain part of the nation's resources to the production of what appear to be useless commodities is not irrevocable; the capacity of a nation to restrict its production of non-necessaries and increase the \'olume of necessaries depends largely upon the proportion of resources devoted to the former. It would be a highly misleading picture of comparative productivity of the two economic systems to compare only the output of necessary commodities and services in such countries as the United States and Soviet Russia. T h e third group of problems in the determination of the specific contents of national income arises from the fact that during normal times, when a consideration of the future is of importance, the performance of the economic system is gauged not by the gross but by the net product, i.e. by the volumes of commodities and services remaining after the replacement of capital outiays; or, in terms of individual and corporate receipts, by the sum of incomes received after subtraction of expenses incurred. T h e most important point in the distinction between g r o s s and

NATIONAL INCOME

17

net is the contrast in procedure between property and labor incomes. I n property incomes the net part is obtained after all capital outlays have been replaced and is thus a pure surplus. I n labor incomes the procedure is quite different. Wage earners or salaried workers consume a part of their capital, viz. theit working capacity, in the process of earning their income; and the replacement of this earning capacity can be accomplished only by consumption at a certain level of subsistence plus provision for complete replacement at the time when working capacity has dwindled to zero. Were we to proceed as in the case of property incomes, we should deduct from wages or salaries the living expenses plus p r o vision for the future and count as net income only the residue, a procedure suggested by some economists (Loria). T h e a c t u a l practise, as is well known, differs materially from that suggested. T r u e , in certain occupations specific professional expenses are deducted from gross revenue in the computation of net income; also the income tax laws, by setting exemption limits and by imposing lower rates on ' e a r n e d ' ' incomes, recognize that a certain part of labor income is not net. But by and large, in contrast to property revenues, in labor incomes the net is almost equal to the gross revenue received. This implies that the working capacity of individuals cannot be treated as a part of the property system; that therefore an ouday by an individual of his personal activity is a part of his general existence and the income he receives is a fund of subsistence and not a mean.s of perpetuating the individual as a part of .•society's wealth. Hence distinguishing gross a n d n e t income of individuals from labor or personal services is a case of drawing a line between the area of the economic principle and the ways of life at large. As this line is drawn, so arc the questions raised solved. For instance, what specific expenses should b e allowed in arriving at an individual's net labor income^ T h e r e is no precise answer, the general basis of determination being the importance to the individual's whole life of the object obtained as a result of expenses. If this importance is great, the ouday, like expenditures on living, cannot be imputed to the income getting activity proper and should not be subtracted. If, however, these

18

CONCEPT OF INCOME AND DISTRIBUTION

expenses arc bound up only with the income getting activity, they can and should be deducted. An interesting illustration is provided by the expenses of doctors, lawyers and other professional persons for their offices or tools, the outlays on which are obviously deductible, and for their education, the expenses for which are usually not deducted. Similar questions arise in connection with wage and salary earners in regard to such cases as the differential expenses incurred in order to live in proximity to place of work. T h e segregation of the purely economic motives in expenses from the broad drives of h u m a n life is a difficult problem whose solution shifts from time to time and country to country. In the United States as in many other countries there is a tendency in the direction of extending the area of expenses allowable for deduction, as the result of a desire on the part of individuals to limit the base of taxable income. T h e same tendency to regard individuals more and more in the nature of capital is illustrated by the legal cases of compensation for injury. T h e exclusion of such payments as industrial compensation for injury from net income is inconsistent with die failure to allow for living and conservation expenses; this is also true of the exclusion of net returns (after subtraction of the past contribution of the recipient) from unemployment, sickness and other social types of insurance. For this reason statistical estimates tend to and should include industrial compensation for injury a t the same time that they exclude insurance compensation for the destruction of a building by fire. Since net income from property can be ascertained only by allowing for the restitution of property outlay, a question arises as to what changes in property should be included in income in order to keep property intact. Shifts in the value of property due to general causes may either be accidental in chairacter, i.e. desQ-uction of property by an earthquake or a sudden rise m the value of property which has survived such a calamity, or may stem from general changes in business conditions, usually manifesting themselves through a rise or fall in the general level of prices. Since national income is to be conceived as the end product of economic activity, measured as net addition of commodities a n d services,

NATIONAL INCOME

19

changes in property values arising from either of these two groups of causes are to be disregarded. But if the changes in property value are a reflection of changes in the surplus of gross revenue over expenses inctorred plus a depreciation charge at secular rates, to include them with the change in such net income would obviously amount to double counting. T h e same objection would hold if the changes in property value are a reflection not of actual rise or decline in net income but of a forecast income shift. It is therefore only in the cases of changes in property value due to plowing back of income or actual impairment of property that, by definition, there takes place a conversion of income into property or property into income; it is only in such cases that net income cannot be confined to the surplus of gross revenue over expenses actually incurred plus a standard depreciation charge. T h e consideration of tlie element of realization of changes in property value does not alter these conclusions. If a property has grown in value because of a general change in the level of prices and the property owner sells it, thus realizing a monetary gain, such gains are still to be excluded from national income; and if they are included in the estimate of national income in current money units, they ought to disappear in the defladon of these estimates by a properly constructed index of prices in which the prices of property are included. T h e consideration of the actual sale of property, however, suggests an additional source of changes in the value of property: an opening u p of better marketing opportunities, due to the professional skill of the property sellers. Such changes take place in cases when buying and selling of property become a professional occupation; and in such cases incomes derived from the sale of property, i.e. from the change in the value of property because of the more skillful handling of it in the market, are to be counted as bona fide net incomes. But this is only another case of the plowing back into the property of certain currendy produced services. In actual statistical practise the inclusion or exclusion of such changes from net income from property is largely conditioned by die prevailing accounting procedure. For example, in a number

20

CONCEPT OF INCOME AND DISTRIBUTION

of extractive industries the depletion of natural resources actually occurring is insufficiently taken care of by the accounting depletion rates, and in such cases net income as recorded by statistical practise contains large elements of gross income. In other industries deductions from current income for depreciation and obsolescence may exceed appreciably the actual destruction of property in the process of production; and it should be noted that the usual practise of relating depreciation and obsolescence charges to the original cost of equipment means, in the prevailing conditions of technical progress, a conservation of capital not at a constant but at a rising productive capacity. T h e practise of reporting inventories at cost or market, whichever is lower, means that in years of declining commodity prices the business units understate their ntet income by deducting losses on inv^entories. A correction in all such cases is rather difficult; and consequently the distinction between property and income, while theoretically feasible, is in actual statistical practise a reflection of the distinction m a d e by current accounting procedure. Of the numerous questions raised above in regard to the contents of national income some can be and have been answered unequivocally by the consensus of learned opinion; others ^re still in the zone of disputation; and still others, while yielding clear answers on theoretical grounds, fail of application because relevant data are lacking. Elements which are generally included in national income received are: wages, salaries, dividends, interest, net rents and royalties, net money receipts by entrepreneurs and independent providers of personal services—all flowing from legitimate pursuits; also commodity receipts of farmers and other selfcontained producers; also the perquisites (such as food and board) of employees and receipts in kind of rent or interest. In order to pass from income realized by individuals to income produced the business savings of the individual and corporate business units and of social units should be added. Items which m a y still be considered in the doubtful zone are: free incomes to individuals from the business system, pensions, compensation for injury, returns from social insurance, hobby products a n d returns from incidental

NATIONAL INCOME

21

services, net services from durable goods used by owners, interest on government loans, the allowance for individuals' deductible expenses and, at a considerable remove, payments to government employees. I n some countries, e.g. India, certain estimates exclude payments for all personal services. Finally, for the more deeply going problem of readjusting the market valuations practise lags far behind theory, even in the seemingly siniple problem of correcting for changes in the value of the monetary unit. National income estimates for various countries m a y differ not only because of the various ways in which the questions concerning doubtful items are answered but also because of differences in the statistical methods employed in arriving at such estimates. These methods vary in their turn because of differences in available d a t a . Such variation is perhaps greatest where the scarcity of statistical materials compels the application of a n arbitrarily chosen coefficient to a statistical measure of only a small part of the total universe. Estimating procedures of this type offer no purely statistical p r o b lems and are usually unreliable, except when undertaken as an extrapolation based upon a long a n d extensive series of income estimates for other years or other areas. Even where statistics cover a substantial proportion of the field, data are rarely available on all types of income yielding acdvity and do not always provide an undistorted picture of income flow. Estimating methods m a y be distinguished according to the type of basic d a t a used: production and trade statistics, income statistics and d a t a on consumption a n d savings. T h e commodity-service method attempts to measure n a t i o n a l , income as the net value of commodities and services produced, tracing them to their originating point in the industrial system. It utilizes the large body of production and trade statistics available in most of the principal countries of the world. Such d a t a permit a comparatively easy discrimination between payments and receipts that represent economic goods and services and those that do not reflect actual creation of new economic goods; for it is the tendency of producdon and trade stadstics to cover only those branches of activity in which the genuinely economic character of

22

CONCEPT OF INCOME AND D B T R I B t r n O N

the activity is firmly established. Often such figures afford also a n approximation to the physical volume of incomes. The disadvantage of this method is the difficulty of guarding against duplications and omissions. In production and trade statistics the full value of a good is usually recorded at each stage of its productive transformation from the point of origin to the sale of the final product. Even when " n e t value a d d e d " is segregated, as is the case in statistics for manufacturing in the United States, a large element of duplication still remains. O n the other hand, production and trade statistics are by their very nature better suited for the coverage of commodities than of services and, among commodities, for the measurement of basic materials of uniform quality than of finished goods of vzu-ying grades a n d brands; there is thus the danger of overlooking certain types of income yielding economic activities. • T h e incomes-received method derives the national total as a sum of net incomes received by individuals and business enterprises. It relies primarily upon the large volume of d a t a gathered by income tax authorities and sometimes upon special studies relating to earnings of various occupational groups a n d to their family budgets. At least in one country, Australia, data were obtained directly by means of a census. T h e method of incomes received escapes the danger of duplications; and the resulting total, in so far as it is based largely on incomes of individuals, is consistent with the concept of national income as it is generally understood in a business economy. T h e main disadvantage of the method is the deficiency in coverage. All existing taxation systems exempt incomes below a certain range or of a certain type, and m a n y of such non-taxable incomes are not even recorded; in the United States, for instance, more than half of the estimated national income is not reflected in income tax statistics. T h e deficiency in coverage becomes less important as the exemption area narrows a n d disappears when a n income census is taken. But census taking is a n expensive procedure and an income census develops defects of its own; the Australian census of 1915 was shown upon analysis to include a n u m b e r of petty receipts whose economic n a t u r e was uncertain. Even for the recorded incomes, tax (and income

NATIONAL mCOME

23

census) data do i. c provide a consistent quantitative picture. T h e statutory definition of income and the character of data requested vary from country to country and time to time. So does evasion of the tax by complete failure to report or by under-reporting incomes received. Moreover the character and statisdcal reliability of income tax data are not easily tested; these are available only in the form of summaries published by government bodies, to which they are merely a by-product of administrative activity. T h e consumption-savings method, which registers income as it flows out of the individual economy, is used less extensively than the other methods, because it calls for d a t a not generally available. No country has as yet continuous and reliable series on the volume of consumers' expenditures and savings, nor even such approxim.ations as would be provided by data on the volume of retail trade or on consumers' budgets at various income levels. Such figures become available either through a greater development of trade, service and banking statistics, in which case the commodity-service method off'ers an easier way of arriving at total national income; or through a further study of the activity of individual households, which is both costly and unpractical because of resistance offered to the inquisitive statistician. At present the consumption-savings approach is used as a stop gap when industrial or income statistics are badly lacking; and data are available on individual savings, on apparent consumption of a number of consumers' goods and to some extent on household budgets. In the future, however, this method may come into greater prominence, for it is increasingly appreciated that a study of the various ways in which income is spent or saved is an essential aid in dealing with a n u m b e r of pressing economic problems. There is an obvious relation between the methods described above and national income at one or another stage of its circulation, but the correspondence is one of practical convenience rather t h a n of logical necessity. Since the different stages of national income are closely related, it may be approached at any stage, that of income produced or received or consimied, and the resulting estimate adjusted for the succeeding or preceding stage. Therefore it

24

CONCEPT OF INCOME AND DISTRIBUTION

is theoretically possible to measure the magnitude of income by any of the three methods suggested; the results should check, and such a check is most desirable. Practically, however, the more compelling rea.son for using more than one method is that because of paucity of data no single type of statistics is sufficient by itself for the purpose of arriving at a reliable national total, unless the investigator gives rein to imagination by employing "raising" or "correction" factors. Estimates utilizing several methods and types of data may be theoretically deficient, because the different constituent parts of the income measure are liable to errors of differing character and hence are not strictly comparable or addible; but the possible error thus involved would be m u c h .smaller than that due to a restriction of the estimate to one type of data with consequent extrapolation over a large field. T h e allocation of national income by different categories is suggested by the questions formulated above as to the specific contents of the total; it arises in the statistical process of building up the total from estimates of its various component parts; and it is necessitated by the recognition that the mere total is not sufficiently illuminating for any of the purposes served by the concept a n d measurement. There is an obvious need for a breakdown of a national income estimate when the total refers to an economic system with widely differing regions; when the industrial constituents of the productive system change; when various types of Income shift in importance; when the form of economic organization changes; and when the personal shares in the national total are unequal to a varying degree. The main types of allocation are: by economic regions, by industrial sources, by forms of economic organizations, by personal distribution according to size of income. T h e need for a regional distribution of income totals has already been suggested. T h e problems raised by such a distribution are similar to those for the national total, but the gravity of some questions is increased in the smaller area units. For incomes conditioned by personal activity of recipients regionahallocation may not b e difficult, unless there is migration of such recipients across

NATIONAL INCOME

25

regional boundaries. M u c h graver problems appear in connection with property incomes. An enterprise may have its plant in one region, sell its products to the country as a whole and pay its dividends and interest to individuals residing in various regions of the country. A regional allocation of property incomes produced or paid out by such an enterprise offers obvious difficulties, problems which have often been discussed by the courts in connection with state taxation of corporations. U p to the present regional allocation of national income has been carried through only to a limited extent. It requires a volume of d a t a not easily available a n d a clarity of concepts so far not attained in the existing income literature. But the increasing interest in regional similarities a n d diversities of economic life may lead to a wider employment of such distributions. T h e distinction among the industrial sources of national income is important because of differences among these fields of activity in the character of work or life for the people employed; in stability, either secular or cyclical, of net incomes derived from them; in the importance of these industries in the general scale of h u m a n wants. Such distributions carried through for one country at successive dates or for various countries at the same date serve to indicate changes in a country's industrial structure over a period of time or structural differences between countries. The distribution presented in Table II for the United States illustrates the type of observation that is facilitated by such statistical measurements. T h e dates at which various trends, such as those from agriculture to industry or trade, become observable differ from country to country and occiir earlier in older nations, such as the United Kingdom or France, than in newer nations, such as the U n i t e d States or -Australia. Similarly, the exact rate at which shifts in the industrial sources take place in the different countries will vary because of differences in a host of natural and institutional determining factors, such as the availability of natural resources, the class distribution of incomes and the extent of purposive control exercised by society. But with ail such differences the decline in the part of agriculture in the nations' end product; the rise, at first

26

CONCEPT OF INCOME AND DISTRIBUTION

TABLE I I .

D I S T R I B U T I O N O F N A T I O N A L INCOME I N T H E U N I T E D S T A T E S B Y i N D X i t m i A L SOURCES*

Percentag es of Total Total Iru-ome {in }1,000,000)

Tear or Period

Agriculture

Mining

Manufacturing

Transportation and Public Utilities

Service and Trade

Gooemr mtnt

1850

2,178

3

35

1

1

1

19

8

19

0

22

0

3.0

1860

3,596

7

30

3

1

7

22

1

20

1

22

4

3.4

1870

6,646

2

26

8

2

2

24

1

11

3

30

2

5.5

1880

7,343

8

20

1

3

0

24

4

12

7

34

1

5.6

1890

11,965

5

18

9

2

7

31

9

10

1

30

8

5.6

1900

17,417

7

21

2

3

4

29

2

9

3

31

6

5.3

1910

29,243

9

23

4

3

3

28

2

9

7

30

9

4.5

1910

29,805

0

19

2

3

2

28

7

10

2

33

9

4.8

1913-17

38,610

0

18

3

3

5

29

8

9

6

33

6

5.2

1913-17

36,652

0

17

2

3

5

28

0

9

4

36

2

5.6

1918-22

58,401

0

16

7

3

4

29

6

10

3

31

7

8.3

1923-27

71,891

0

11

1

3

2

27

8

9

5

40

5

7.2

19301

72,141

0

8

0

2

3

26

4

10

0

44

3

9

0

• T h e Ihrcc d i v i s i o n s o f t h e table r e p r e s e n t t h r e e diffTcnt e s t i m a t e s w h i c h a r c n o t s t r i c t l y c o n n p a r a b l e ; t h e d e g r e e o f d i i p a n t y b e t w e e n t h e m m a y b e j u d g e d f r o m the t w o sets o f figures for 1 9 1 0 a n d 1 9 1 3 - 1 7 .

Tbe

first a n d l e c o n d p a r t s o l t h e t a b l e — t h e figxircs f o r 1 8 5 0 - 0 ; D a n d for 1 9 1 0 t o 1 9 1 3 - 1 7 — r e p r e s e n t total i n c o m e includiDg b u s i n e s s s a v i n g s but e x c l u d i n g g o v e r n m e n t rent a n d i a t e r e s t a n d m i s c e l l a n e o u s i n c o m e ; t h e t h i r d p a r t — t h e figures f o r 1 9 1 3 - 1 7 t o 1 9 3 0 — r e p r e s e n t s r e a l i z e d i n c o m e e x c l u s i v e of g o v e r n m e n t r e n t and i n t e r e s t » n d mist:c'ilancous i n c o m e . fiveryear

T h e figurci f o r 1 9 1 3 - 1 7 to 1 9 2 3 - 2 7 a r c a n n u a l a v e r a g e s f o r t h e c o r r e « p o o d i n g

periods

t Preliminary Egures. Source: K i n g , W . I., Tfu Width

and Tncome of the Ptople of tht United States

o f E c o i w i a i c R e s e a r c h , rncome in ths Uniud Matioaat

Income and Iti Purcharing

Pouf

StaUs,

( N e w Yoik 1919); National Bureau

2 v o l s . ( N e w Y o r k 1 9 2 1 - 2 2 ) v o l . ii, pt. i , K i n g , W . I . , Tht

National Bureau of Economic Research, Publication no. 15 ( N e w

York 1 9 3 0 ) ; a n unpubliahed estimate by t h e N a d o n a i Bureau of E c o n o m i c R e i e a r c h o f n a a o n a l iiscomc for 1930.

rapid and then disappearing, in the relative contribution of mining and manufacturing; a n d the increase, especially marked in the United States during recent decades, of the share coming from service, trade, finance a n d governmentj are tendencies which appear as constituent elements of the growing capitalistic system of production and thus characterize all countries drawn into the path of its evolution. T h e figures resulting from the allocation of

NATIONAL INCOME

27

national income by industrial sources are thus measures which afford specific confirmation of the broad tendencies of industrial evolution observed otherwise by historians and economists; or of the broad differences, generally known, in the industrial constitution of various countries. T h e net contribution of such measurements lies only in refining a n d checking the generally held notions on these subjects. And when available annually they may aid in the study of the relative stability of income flows from various industrial sources, a problem on which the prevalent generalizations still need considerable testing. There is a natural tendency to identify the industrial groupings, which are institutional categories, with the more analytically derived types of economic activity (extractive production, manufacturing production, distributive trading, finance, transportation and so on) and to infer that a shift in the relative importance of a given industrial source, such as personal service, trade and finance implies an identical shift in the extent of trading, financing and personal service activity. This, however, is not necessarily the case, for with changes or differences in the social division of labor the exact scope of activities subsumed under an identical industrial group may change or differ considerably. T h e manufacturers of the United States may have been distributing, financing and providing personal service to a greater extent (relatively to their purely manufacturing activity) several decades ago than they are doing now. T h e increase in the relative contribution to the national total shown by finance, trade and services m a y therefore be due partly to a shift of financing, distributing and service functions from manufacturing and other activities to a separate professional group. A similar lack of identity between industrial groupings and types of economic activity affects comparisons among vcirious countries. Clearly the difficulty of inferences from allocations of national income by industrial sources is the greater the more specific such allocations are—even neglecting the fact that carrying industrial distinctions beyond a few major groups runs afoul of the absence of definite criteria as to what constitutes an industry or an industrial group.

28

CONCEPT OF INCOME AND DISTRIBUTION

It may be said that the general evolution of the industrial system is toward a more intensive division of labor, a greater specialization of functions and hence a closer identity of industrial groupings with types of economic activity; and that consequently the shifts in the relative shares of industrial sources tend to result in overestimation of the shifts in the relative importance of production versus transportation, transportation versus distribution and so forth. But such a general inference neglects two difficulties. T h e first is the existence of a counteracting tendency of vertical integration which complicates the proper allocation of single economic units; this difficulty may be enhanced by special factors, such as the allowance in the United States of consolidated income tax returns from corporations. In the second place, the difficulty of determining how far the division of labor has gone in the direction of segregating a new industrial division is especially disturbing, the best illustration being provided by the treatment of interest on loaned funds. If a bank receives interest payments on short term credits to a pig iron manufacturer, is this income produced by the banks or by the pig iron industry? T h e current statistical practise con.siders such incomes to be produced by the banking industry. But do banks produce the interest received by them on government securities? Are interest payments received by an individual on his railroad bond income produced by the individual or by the railroad? T h e current practise is to consider these interest payments as income produced or paid out by the government or the railroads, the basis of such decisions being the distinction between professional activity requiring skill and experience, as exemplified in a bank's commercial credit policy, and a non-professional activity of investment requiring no such qualifications. But the carrying through of such a distinction is beset by difficulties. A r e the incomes of holding companies, insurance companies, savings banks, investment trusts and similar institutions, whose main source of income is equities in other concerns, to be counted as the product of insurance or tlie investment industry as such, or are they to be allocated to the industrial activities which constitute their primary origin? Such questions have arisen but seldom in statistical practise, mainly

NATIONAL INCOME

29

because o f lack o f d a t a , but they will have to be faced in the n e a r future. For all these reasons the apparently precise results of distributions of national income totals by industrial sources must b e interpreted with a great deal of caution, being an approximate reflection of only the broadest trends or differences, and need to b e supplemented by allocations of the national total based on other criteria. Forms of economic organization may be distinguished by types of the organized unit (individuals and corporations), by the general principle of organization (free competition, regulation and complete control) or by any of a number of basic elements. N o matter how the form of economic organization is defined, the national economy of the last century and of the recent decade? represents a combination of branches functioning under different forms o f organizations. T h e distinction of the relative importance of the latter on the basis of shares of national income derived from activities organized upon different ruling principles is of considerable significance. But the difficulty of such distmction lies in the contrast between the absolute categories set up by analysis and the absence of such pure forms in reality. O n e could presumably formulate adequate definitions of free competition or of complete control; but to establish whether or not a given branch of activity is in a state of free competition is difficult, even with access to the internal records of the individual enterprises in the field. O n the other hand, when the possibility of a clear distinction is given by formal criteria, e.g. corporate and non-corporate units, such criteria may not correspond to the essential meaning of the distinction. From the point of view of economic analysis the one-man corporations, which formally belong to the corporate field, are by the nature of their operation m u c h more similar to individual businesses than to the giant, anonymous corporate units. It is this difficulty that largely explains the failure of statistical study of national income to pay proper attention to the allocation by forms of organization. Such a breakdown takes place mostly m s o far a s i t is coincident with allocation by industrial sources, a

30

CONCEPT OF INCOME AND DISTRIBUTION

coincidence t h a t is relevant since the technological diflFerences among industrial divisions form a basic element which underlies differences in form of organization. T h u s the estimates presented in Table II permit one to draw inferences, from the relative growth of such corporation dominated industries as mining, manufacturing, transportation and finance, as to the growing share of national income paid out by corporations; and from the increasing share of government and public utilities in the total to draw inferences as to the growth of controlled areas of our economic system. Further precision in the distribution by forms of organization is at present impossible because of lack of data. But since such breakdowns are especially important in a national economy of a transitional type, when changes in organization are rapid a n d their effect has to be measured as a basis of economic prognosis or diagnosis, and since recent developments have stimulated changes in the relative areas of various principles of organization, the near future is likely to witness an increasing emphasis upon the allocation of national income by forms of organization. The distribution of national income by forms of payment is a n attempt to go beyond the industrial and organization groupings and to measure the current returns of such general productive factors as labor, capital and land. T h e significant political and social conflicts that center about the relative share of these productive factors render a quantitative measurement and test supremely important. An illustration of results obtained by such measurements for the United States is provided in Table I I I . T h e difficulty of obtaining consistent estimates and hence of arriving at definite conclusions is shown in this table by the twc sets of ratios for the year 1910. Such inconsistencies make a comparison of the distribution for various countries impossible without a thorough reanalysis of the published data and some rather arbitrary adjustments. But the broad trends in the United Stateover a period of time, as shown by T a b l e I I I , can be said to be fairly typical of other industrial countries. Wages and salaricappear to account for a slightly rising proportion of the nation?! total; while entrepreneurial income, which is a combination of

31

NATIONAL INCOME TABLS I I I .

D I S T R I B U T I O N O F N A T I O N A L I N C O M E IN T H E U N I T E D S T A T E S B Y F U M C T I O N A I . SOURCES *

Percentages of Total

Tear

Service Income

Total Income (in \sipoopoo)

Property Incortu

Wages and\ Salaries

Enttrprenturial Income 44.7 39.8 31.9 21.4 24.8 30.9 28.9

77.4 81.1 73.2 78.8 79.6 77.7

32.4t 32,Of 29.4t 26.31 24.7t

89.9t 89.Sf 92,3t 91.8t 89.8t

1850 1860 1870 1880 1890 1900 1910

2,178.3 3,596.7 6,646.2 7,343.8 11,965.5 17,417,7 29,243.9

36.4 37.6 49.2 51.8 54.0 48.7 48.9

1910

29,805.0

55.6

1910 1913-17 1918-22 1923-27 1928

28,297.0 36,652.0 58,401.0 71,891 . 0 78,502.0

57.5 57,5 62.9 65.5 65.1

Total

Rent

81.1

7.3 8.6 6.6 8.5 7.3 7.0 7.7

Interest and Dividends 11.6 14.1 12.3 18.3 13.9 13.4 14.7

Total

18.9 22.7 18.9 26.8 21.2 20.4 22.4

10.1 10.5 7.7 8,2 10.2

* T h e t h r e e d i v i s i o n s o f t h e t a b l e r e p r e s e n t t h r e e d i f f e r e n t e s t i m a t e s , t h e d e g r e e o f d i s p a r i t y b' upon a piecing together of the most variegated and detailed evidence, since the data do not permit a brief summary such as is provided in national income estimates for later years. Such summary estimates, however, must be interpreted with the utmost degree of caution, if misleading inferences are to be avoided. Just as in the allocation of national income by industrial sources so also here there is a nattiral tendency to identify institutionally determined divisions with analytical categories. Wages and salaries tend to be identified with the theoretical category of labor income; and conclusions are often d r a w n as to the increase in Inhor's share o f the national product from the rising percentage of wages and salaries in the total income. But such an inference is obviously misleading because of: first, the increasing relative weight o f industrial branches in which the corporate form of organization predominates ( e g . the share of manufacturing increases while that of agriculture declines); and, secondly, the increasing weight within each indusUy of the corporate form of organization. As a result of both tendencies the share of labor payments, which has formerly been combined with other functional pavmenls in the mixed category of entrepreneurial incomes, is increasingly segregated and goes to swell the relative weight of \^•ages say somewhat plaintively, " y o u come along with the Gross National Product and ci-eate a lot of confusion. What is it needed for? H o w does it differ from the national income and what advantage does it have in practical economic a n d business affairs?" Such questions are answered in this article, describing the concept of the gross national product. Certain technical details have been omitted so as to focus attention on the central ideas involved. As an introduction to both the meaning and uses of the gross national product concept, it is helpful to recall the economic problems of 1941. These presented the immediate occasion for compiling gross product statistics. T h e nation's rearmament program was in process of bein at the disposal of )'our business. It is significant that other countries felt the same need for statistical background material. During its months of greatest peril at the time of the "blitz," tlie Er'tish government set a small group of statisticians to work preparing the first oflicial esdmates of national product a n d nadonal inr .-jme. This was done because it was proving so difficult to organize the economy for war without these guideposts. INCOME AND PRODUCT STATISTICS

T h e statistics on national income were the nearest thing to current infoimation of this type on hand at the inception of our rearmament program. As readers of Charles L. Mcrwin's ardcles

46

CONCEPT OF INCOME AND D B T R l B i m O N

in Dun^s Review (August and November 1942) will recall, national income is the sum of the earnings of the various factors of production for their participation in the producdve process. A table of the national income for any year, say 1942, looks about like this:

T o t a l national income, 1942 Salaries and wages N e t income of proprietors .Agricultural Other business and professional Interest and net rents Corporate profits after taxes Dividends Undistributed profits

Billions $119.8 83.7 20.1 9.7 10.4 8.4 7.6 4.0 3.6

Clearly, this information was relevant for the purchasing power side of the picture. But it was not the whole story and consequently could not, by itself, be brought to bear effectively on the problems under discussion. The first requirement, then, was to develop data on total current production. Since the problems at issue concerned the allocation of output and resources between Government a n d private uses, it was necessary to think of this total production as consisting of the output of the private business system plus the output of Government. A little reflection will show that the output of private business, exclusive of materials and business services used in the process of production, can be grouped according to its three major outlets as: goods and services sold to consumers; total capital goods sold to, or retained by, business; and products sold to Government. When the goods and services produced directly by Government, which most conveniently can be measured by the costs of the productive factors used by Government in their production, are added to the output of business one obtains the desired measure of total production. Gross national product simply is the technical name of this measure of total production. It may be defined as the aggregate value of the current production of goods and services flowing to

NATIONAL PRODUCT AND INCOME STATISTICS

47

the Government, to consumers, and—for purposes of gross capital formation—to business. I n the breakdown in which it is currently published it looks about like the following table: Billums Gross national product or expenditure, 1942 Government production a n d purchases of goods a n d services. War Non-war Private gross capital formation Construction, machinery and e q u i p m e n t N e t change in business inventories Other Consumers' purchases of goods a n d services

$151.7 61.7 49.3 12.4 8.0 8.0 — .6 .6 82.0

It may not be clear immediately why it is permissible to break down a total, which refers to current production, into components that refer to sales or purchases, that is, why gross national product is the same thing as gross national expenditure. T h e explanation lies in the fact that the change in business inventories, the balancing item which constitutes the difTerence between current sales and current production, is included in the total as a component of capital formation. Another feature of the tabulation which requires comment is that capital formation is taken on a gross rather t h a n a net basis— that is to say, replacements of existing equipment and construction as well as net additions to the stock of capital goods are included. T h e reason for this is twofold. In the first place, gross capital formation was easier to estimate than net because of certain technical characteristics of the available sources of data. These center around the fact that the estimates of new capital output must be derived from production d a t a whereas the estimates of capital used u p must be derived from financial records. These two sources do not m a t c h at all well. As the statistics stand, the depreciation estimates cannot be subtracted from the gross capital formation to yield a meaningful estimate of net investment. Although the statistical difficulties probably can and will be solved, there has not been time to solve them as yet.

48

CONCEPT

OF

mCOME AND

DISTRIBUTION

Secondly, attention is focused on gross rather than on net capital formation because for many purposes it is the more important magnitude. When in connection with the war program, for instance, it had to be determined what \ o l u m e of resources could b e freed for war production by diminishing private capital formation, it was gross capital formation that was relevant. For, in the short run, resources could be made available not only by not adding to the stock of private capital, but also by not making certain ordinary replacements. A national product estimate which would have included only net capital formation would have concealed an important potential source of war output, and would have led to exaggerated notions of the necessary curtailment of consumers' goods. In many cases it is a lot easier to struggle along with old machinery and buildings than it is to have less food, clothing, or even amusements. SUMMARY

OF M . ^ J O R

TRANSACTIONS

IN

ECONOMY

Hypothetical figures m billions of dollars I.

Dispusat

RKtipt!

of

( a x e s . . . . 25

taxes

\ . S a l e s of

5. P a y o f f a c -

4, P a y r e -

Receipts

1. P u r c h a s e s o f

ceived

consumers'

business.. . 50

poods and

duction. .100

from Gov-

goods and

. 80

cmmsnt ..15

3. S a l f s of

7.

tors o f p r o -

products

d u c t i o n ....15

to G o v e r n -

taxcj

25

«r\-iccj. . 80 10.

5. P a y r e -

Biuiflc«

Personal taxes

ceived

S

from business

SO

Transfer

100

9. T r a n s f e r

paymcDia. . 5

payments. . 5

1 2 . BoiTowmg

«/

tors o f pro-

ment 9.

Diip^al

Receipts

Receipts

consumer

4. P a y o f f a c 5

of

products of

scrs i c e s . 10. P t n o n a )

Disposal

Receipts

Rictipll

3. Purchases of

7. B u s i n e s s

Indioiduals

Busintss

Gooernrjwit

2. Private 40

6.

Depre-

11.

Indi-

grois

ciation

vidual

capital

and other

vavinga. . 3 5

fomation.. .5

reserves . . . 6 8, U n d i s tributed profits

Total

70

701

1

135;

i

4 135

120

120

NATIONAL PRODUCT AND INCOME STATISTICS

49

RELATIONSHIP ILLUSTRATED

Since the national income is generated in the production of the national product, the reader probably is wondering what the relation of national income to gross national product is, and why the latter is so much larger than the former. It is necessary that this relation be understood if intelligent use is to be m a d e of the statistics. T o clarify this relationship it is convenient to coi^ider separately the two major components of gross national product, namely, Government product and private business output. So far as the direct contribution of the Government to gross national product is concerned, it is measured, as was mentioned earlier, by the costs for factors of production incurred by Government. Hence, it is exactly matched in national income by the income received by factors of production employed by the Government because national income is simply the earnings of all the factors of production utilized in the economy. T h e relation between the rest of gross national product (sales of private business to Government, to consumers, and private gross capital formation) and the rest of national income (earnings of the factors of production employed by private business) can best be seen with the aid of the profit and loss statements of private business. If the profit and loss statements of all business firms are consolidated, current cost item.s that constitute purchases from other firms will cancel against corresponding receipt items in the accounts of the latter. For instance, raw materials sold by firm A to firm B will appear once as receipts in the accounts of firm A and once as a current cost in the accounts of firm B. T h e two entries will cancel out when the profit and loss statements are consolidated. Hence, on the receipt side, there will be left sales to consumers, to Government, and private gross capital formation. But this is exactly the remainder of national product for which we are trying to account in terms of incomes. O n the income and expense side, there will be left payments to factors of production (including dividends), depreciadon and other reserves, taxes, and undistributed profits. These items will exhaust

50

CONCEPT OP INCOME AND DISTRIBUTION

the Ust. For current cost items with respect to purchases from other firms will have been eUminated in the process of consolidation. This accounts for all the income and expense items that correspond to gross national product. Assembling and rearranging them, we see that they equal the earnings of the factors of production as included in the national income with certain additions. These charges against business sales which are not included in the national income are depreciation and other reserves, a n d taxes paid or o v ing by business. Thus, depreciation and similar charges plus business taxes constitute the two additions to national income needed to make it match gross national product. Business taxes are of two main types: taxes that are treated as ordinary expenses in business accounting—principally sales taxes, excise and business property taxes, and so on—and corporation iijcoine and excess profits taxes which are deducted from income to determine the amounts available for dividends and undistributed profits. T h e relation between national income a n d gross national product is illustrated by the following table: Billions National income, 1942 Plus; Business T a x e s Depreciation and other reserves Equals: Gross national product or expenditure, 1 9 4 2 USE

OF

$119.8 24.0 7.9 151.7

INCOME

In tracing the income and expense flows that correspond to gross national product we incidentally have accounted for the manner in which private business as a whole disposes of its receipts since the income and expense side of the consolidated profit and loss account of private business is nothing more than a statement of the disposition of business receipts. A corresponding statement of the accounts of individuals, essendal to the analysis of the inflation problem, now may be developed. To arrive at the income of individuals it is necessary to deduct undistributed corporate profits from national income, a n d to a d d transfer incomes (pensions, net social security payments, a n d relief)

NATIONAL PRODUCT AND INCOME STATISTICS

51

paid by Government to individuals. The latter are excluded from national income on the grounds that they are not received for productive services, b u t they do represent currently received purchasing power to the individuals receiving them. T h e sum total of individual incomes is used p a r d y to buy consumer goods and services, and pairtly to pay personal taxes of various kinds such as income and estate taxes, personal property taxes, and licenses. T h e balance of incomes, not spent on consumption or paid in taxes, must constitute saving. The following table of the disposition of individual income is the result of these calculations. National i n c o m e , 1942 Add: Transfer payments Less: Corporate savings Contributions t o social insurance funds Equals; I n c o m e payments t o individuals Les.s: T a x e s paid by individuals Equals: Disposable i n c o m e of individuals Less: Consumer expenditures Ekjuals: N e t savings of individuals

SillioTU $119.8 2.6 3.6 3 3 115.5 6.6 108.8 82.0 26.9

It may be noted that in developing statements on the receipts of businesses and individuals and their disposition, the d a t a necessary for a similar statement for Government incidentally have been assembled. O n the one hand, there are the major categories of Gk)vernment expenditures—pay of factors of production, piorchases of goods and services from private business, and transfer payments. O n the other hand, there are Government revenues—personal and business taxes. The missing item needed to balance the expenditures and receipts sides of Government accounts is borrowing, or the Government deficit. BIRD'S-E\'E

VIEWPOINT

In the course of constructing national product and income statistics, we have dealt with the basic aggregates required to give a summary view of the economic system in terms of the analytically important types of transactions, such transactions being shown in their interreladon to each other. T o think of the statistics in these

52

CONCEPT OF INCOME AND DISTRIBUTION

terms—as a bird's-eye view of the economic system—is the most fruitful approach that can be taken in making use of national product and income data. T o show clearly that they consdtute such a picture the d a t a may be rearranged as in the summary table accompanying this article. The table, on page 48, has three double columns, one eacli lor Government, business, and individuals, the three groups whos*:' interplay determines the working of the economy. T h e lefthand side fjf each column shows the receipts of each group. T h e right-hand side sliows the m.anner of their disposal. The coinpoaents of gross national product and income may be fitted into tht columns of this tabic. T o facilitate the task they are recapitulated in the order in which they were mentioned. 1. 2. 3. 4. 5.

(Consumer goods and services. Pri\ ate gross capital formation. Products of private bu.siness sold to Government. Pay of factors of prodiiction employed by Government. Pay of factors of production employed by business (including dividcrids). • . 6. DepreciatK'n and oi'ier reserves. 7 . Business tii.xes.

8. 9. 10. 11. 12.

Und-;'.ributcd profits. Transfer payments. Personal taxes. Individual savings. Government borrowing.

Let us firsi enter items for which the interrelation between the three m..; ir accounts is most transparent, that is, those for which an entry w. the leit-hand side of one column is clearly matched by an entr\- in the right-hand side of another. Thus consumer goods and services, item 1, are put in tlie left-hand side of column II as a receipt of business. They also appear in the right-hand side of column i n as an expenditure of individuals. Items 3, 4, 5, 7, 9, and 1 0 are handled similarly. This completes the items to which clear-cut counter-entries correspond.

NATIONAL PRODUCT AND INCOME STATISTICS

53

Next the accounts of the Government are balanced by entering borrowing, 12, in the left-hand side of column I. T h e books o f business are balanced by entering private gross capital formation, 2, in the left-hand side of column II and depreciation and other reserves, 6, and undistributed profits, 8, in the right-hand side. Finally, the accounts of individuals are balanced by entering individual savings, 11, in the right side of column I I I . W e have balanced all the accounts, and fitted into the columns all the components discussed. Examination of this table will show that it includes the aggregates necessary to give a complete summary of the economy in terms of its major transactions. It is not surprising, therefore, that national product and income data form the basic statistical background and point of departure for the study of economic problems which affect the nation as a whole. These statistics also can be looked upon as the first outline, as it were, of a detailed picture of the economy which can be much further refined. For instance, consumer goods and services can be subdivided, showing the types of goods and services bought by consumers. Gross capital formation similarly can be classified. Government expenditures can be broken down, either by the type of products bought by the Government, or by the type of service provided to the community. Income statistics, in turn, can be classified by distr butive shares, or by industrial origin, or by size of total income. Taxes can be grouped in whatever manner seems most useful in the discussion o f tax problems. And savings can be broken down into currency, bank deposits, saving bank accounts, life insurance, bonds, stocks, and so on. Th," filling in of all this information gives a more detailed view o f the economy and considerably enhances its usefulness. But n o matter how the tables m a y be rearranged or w h a t refinement o f detail may be introduced, it is important to recognize clearly that the national income and national product are i n essence neither more nor less than a summarization of the receipts and expenditures sides of the books of business, Government, a n d consumers. This means that the statistics are subject t o the limitations o f

54

CONCEPT OP INCOME AND DISTRIBimON

accounting practice but it also means that the various categories in the tables are essentially those used in, and hence significant for, practical business and economic affairs. . T o have the income and product statistics presented in a set of interrelated tables, instead of having to collect them from a number of independent sources, facilitates quantitative comparisons between the various series. It also is an inestimable aid to clarity of thinking. T h e tables show how the various magnitudes are conceptually related to each other and indicate the legitimate comparisons and operations to which they can be put. T w o examples may be given in this connection. COMMON FALLACIES

Prior to the presentation of the statistics as an interrelated set one of their most frequent misuses was somewhat as follows. Ek;onomists would make a forecast of Gk)vernment expenditures and of national income, in the light of the war program, and would deduct the former from the latter to estimate the amounts available for private capital formation and consumer expenditures. T h e statistics as now presented should guard their tisers from this pitiall. They show clearly that Government expenditures (including transfer payments), consumer expenditures, and private gross capital formation add up, not to national income, b u t to national income plus transfer payments plus depreciation a n d depletion charges plus business taxes. Hence deduction of Government expenditures from national income to estimate the goods a n d services available for private use is an error which yields m u c h too low a figure. More subde misuse of the statistics is m a d e in connection with the savings data. It recurs in infinite variations and constitutes one of the leading fallacies in the interpretation of the statistics. Only the variant that is most important at the present will be mentioned. In discussions of fiscal policy the high level of savings often is adduced as proof that the danger of inflation is vastly overrated. For, it is argued, the Government deficit is matched

NATIONAL PRODUCT AND INCOME STATISTICS

55

by a huge volume of savings, so that the net upward pressure on prices is negligible. SAVINGS AND INFLATION

This argument is based upon a complete misconception of the nature of the statistics. As statistically measured, savings in excess of private gross capital formation always equal Government borrowing whether there is an inflation going on or not. T h i s is revealed by an examination of the statistics as a n interrelated set. Turning to the summary table, one sees that all the items above the first horizontal line cancel out. This is so, because they consist of a set of double entries. We also know that the column totals below the second horizontal line cancel out, since these were derived by balancing the left- and right-hand sides of the columns. Accordingly it follows that the items between the horizontal lines must be equal, that is, Government borrowing plus private gross capital formation equals individual savings plus undistributed profits plus depreciation and other reserves. Since the summary table was derived without any assumptions regarding the presence or absence of a n inflationary process, it follows that the above relation holds under all circumstances, a n d that it does not indicate a state of balance in the economic system. This proof may leave the reader uneasy. " W h a t , " he m a y ask, "would happen if sufficient savings are not available a n d the Government prints money or borrows from the banks to cover the deficit? Will not the deficit exceed savings?" Suppose that in the situation depicted in the summary table the Government prints $10 billion of additional money and spends it on the products of private business. The Government deficit then will have increased by $10 billion. But the same $10 billion also appears as the receipts of private business and, provided that there is no change in other items, the undistributed profits of private business will rise by the same amount. Total savings will have increased exacdy in the amount of the deficit, and the statistics will not indicate that new money has been created.

56

CONCEPT OF INCOME AND DISTRIBUTION

Suppose thin the Government did not obtain more in goods from private business than it obtained prior to the increase in its expenditures, that is, that the additional $ 1 0 bilhon merely went to sw